Doctors spend too long on their computers trying to accurately determine the correct icd10 and HCC code that describe their patients health conditions.
What it does
We use natural language processing and AI to automatically interpret a doctor’s free text notes and then provide suggested ICD10 codes that accurately describe the patient health conditions, along with detecting symptoms and estimating potential diagnoses.
How we built it
We created and designed the web application with a MERN stack (MongoDB, Express, React, and Node.js) along with redux on the frontend, and we created an NLP script on the backend that extrapolates key words from doctor's notes in java.
Challenges we ran into
We ran into problems with given time restrictions, creating a working NLP script, and designing a proper UI to meet the needs of our prospective customers.
Accomplishments that we're proud of
We have key word detections using NLP for given doctor's notes. We also make use of the APIMedic API to estimate potential diagnoses from a given list of symptoms
What we learned
We learned quite a bit about potential issues that doctor's face every day, along with a deeper understanding of the patient doctor workflow that comprises the U.S. healthcare system.
What's next for DiagnosisIQ
We hope to gather larger amounts of data in order to allow for a comprehensive experience that has no gaps in information.